Objective: Lung cancer management has changed due to emergence of new imaging techniques and of multimodal therapies. Our purpose was to analyse how lung cancer evolved in surgical practice.
Methods: The records of patients who underwent surgical resection for lung cancer from 1983 to 2006 in two centres were reviewed. Data were split into four time periods of 6 years. We analysed and compared the epidemiological, pathological and prognostic characteristics of each period.
Results: There were 832, 1148, 1493 and 1195 patients during the periods 1983-88, 1989-94, 1995-2000 and 2001-06, respectively. The main changed characteristics were increasing numbers of older patients, females, past history of another cancer and/or cardio-vascular disease, adenocarcinomas and undifferentiated large-cell carcinomas, smaller tumour size, T1-T2, N0 (47.2-61.2%) and neoadjuvant therapy (NAT) (3.8-24.9%). There were also a decreasing number of exploratory thoracotomies, pneumonectomies and adjuvant therapy (AT) (48.5-30%). The 5-year survival rates improved (34.5-46.3%, P < 10(-6)), mainly after lobectomy, and in the case of adenocarcinoma, N0 and N2 patients. Multivariate analysis confirmed that time trend was an independent factor of prognosis (P < 10(-6)), just as important as N involvement, complete resection (R0), tumour size, age, another cancer history and more significant than the type of resection, histology, NAT and AT.
Conclusions: During the last 25 years, the clinico-pathological features of operated patients have progressively changed and the results following surgery improved. Earlier stage diagnosis might explain overall survival improvement, and play a more major role than associated peri-operative treatments. Therefore, it is advisable to consider the time-related factor in future studies on lung cancer surgery.